|
 |
Deleted files: chances for recovery |
Depending on a file system, OS does different actions to complete file deletion. As for Windows FAT file system it marks file directory entries as 'unused' and destroys file allocation information (except file start), regarding NTFS it just marks file entry as unused, deletes record from directory and marks disk space as unused; as for most Linux/Unix file systems they destroy file descriptor (the information about file location, file type, file size etc.) and marks disk space as free.
The common action for each deletion procedure is to release storage space used by file. This storage space is not wiped immediately (for performance reasons), thus actual file data remains on disk until this storage space is reused to store a new file.
|
|
|
|
Windows file systems
The main Windows file systems are FAT (with FAT32 extension) and NTFS. The more recent ExFAT file system, the completely new implementation of FAT file system idea is not widely used yet. Like any deleted data recovery, the file recovery from these file systems is possible until actual disk space is re-used for a new file.
File system | File deletion action | Non-fragmented file recovery | Fragmented file recovery | Recommended to recover files with |
| FAT | Directory record marked as 'unused'. Clusters are marked as 'free' that destroys chain of clusters, used by file. | File name, size and on-disk position remain inside of directory record so file recovery chances are near 100%. | The chain of file clusters is destroyed so there is no information about file fragments. The are known file name, file size and start position. Using heuristics it's possible to guess fragments locations but there are no guarantees for correct guess. | Raise Data Recovery for FAT UFS Explorer Standard Recovery (version 4) |
| NTFS | Master File Table record marked as 'unused'. Bitmap of used space is updated to release used clusters. File entry is deleted from directory record. | File name, size and on-disk position remain inside of Master File Table record so file recovery chances are near 100%. | File name, size and fragments chain information remains inside of Master File Table record so file recovery chances are near 100%. Lower recovery chances are for very fragmented files which use more than one Master File Table entries. | Raise Data Recovery for NTFS UFS Explorer Standard Recovery UFS Explorer Professional Recovery |
MacOS file systems
This section includes deleted files recovery for HFS+ file system (the old HFS file system is not used anymore):
File system | File deletion action | Non-fragmented file recovery | Fragmented file recovery | Recommended to recover files with |
| HFS+ | It wipes data from B-Tree metadata records for the file and updates map of free space. | File name, size and on-disk position are wiped, however file system journal still may contain this information that allows recovering good files. There are quite high chances to recover lost information with IntelliRAW™, whereas file name information can be lost. | The same as for non-fragmented files, however, if a file is too fragmented, the information about non-resident file fragments is destroyed permanently. | Raise Data Recovery for HFS+ UFS Explorer Standard Recovery (version 4) |
Linux file systems
This section includes analysed XFS, Ext2/Ext3 and Reiser file systems.
File system | File deletion action | Non-fragmented file recovery | Fragmented file recovery | Recommended to recover files with |
| XFS | It clears information part of file node and updates tree of free blocks. Information about file name is unlinked from directory entry. | Using heuristics, it's possible to find file name and rounded to block file size and position. Chances for recovery are near 100%; whereas chances to retrive real file name are near 80%. | File name, size and fragments chain could be retrived using heuristics. If file data is not damaged, file recovery chances are near 100%. Near 80% chances are to assign real file name. | Raise Data Recovery for XFS UFS Explorer Standard Recovery (version 4) |
| Ext2 | It marks file node as free and updates map of free blocks. Information about file name is unlinked from directory entry. File name to node reference is wiped. | Information about file start and size could remain on disk. Nodes analysis can helps to recover good files, however file name information is lost. | The same as for non-fragmented files. | Raise Data Recovery for Ext2/Ext3 UFS Explorer Standard Recovery (version 4) |
| Ext3 and Ext4 | It wipes file node and updates map of free blocks. Information about file name is unlinked from directory entry but still references right node. | Information about file start and size is destroyed permanently however may remain in file system journal. There is no link between file name and on-disk location. Using heuristics and journal analysis it is possible to recover good files even with real names. | Usually, there is no information about first 12 blocks of a file. There is also no information about file name and size. Chances for deleted file recovery are quite poor, however information about most recently deleted files may still remain in file system journal and there could be 100% recovery chance of a file with real file name. | Raise Data Recovery for Ext2/Ext3 UFS Explorer Standard Recovery (version 4) |
| ReiserFS | It updates its S+-tree to exclude a file and updates map of free space. | The S+-tree node still may remain on disk (the copy in file system journal and old copy, created with copy-on-write). In this case file recovery chances could be up to 100% | The same as for non-fragmented files. | Raise Data Recovery for ReiserFS UFS Explorer Standard Recovery (version 4) |
| JFS (JFS2) | It updates object use counter and releases inode in inode use map. Directory is rebuilt to reflect changes. | File inode still resides on disk so file recovery quality is about 100%. Low chances for file name recovery. | The same as for non-fragmented files. | Raise Data Recovery for JFS UFS Explorer Standard Recovery (version 4) |
BSD, Solaris, Unix file systems
This section includes only analyzed UFS and UFS2 file systems.
File system | File deletion action | Non-fragmented file recovery | Fragmented file recovery | Recommended to recover files with |
UFS UFS2 | It clears file node and updates map of free blocks. Information about file name is unlinked from directory entry. | Information about file start and size is destroyed permanently. There is no link between file name and on-disk location. Using heuristics it is possible to recover good files of known file type, however non-fragmented files rarely occur on UFS due to specifics of its Soft Updates algorithm. | There is no information about first 12 blocks of a file. There is also no information about file name and size. Chances for deleted file recovery are quite poor; however, it is still possible. | Raise Data Recovery for UFS UFS Explorer Standard Recovery (version 4) |
Clustered file systems
Data recovery from file systems like Apple Xsan (CentraVision file system, StorNext file system), RedHat Linux Global File System (GFS), VMware ESX Server Virtual Machine File System (VMFS) and so on has been analyzed by LLC SysDev Laboratories but will not be available in our retail software products. If you are interested in data recovery from these file systems - feel free to use our remote recovery service or bring such systems to laboratories of our partners.
|
| Copyright © 2004-2010 SysDevSoftware, the Development & Research division of SysDev Laboratories LLC. All rights reserved. |